论文部分内容阅读
模糊马尔可夫随机场在多值图像分割问题中多采用分段模糊的方法将多值问题转化成多个两值模糊问题,这种方法的模糊度完全依赖像素的灰度值,从而很容易陷于局部最优解。基于可能性测度的模糊随机场摆脱了隶属度对灰度的依赖,使分割结果更容易收敛于全局最优。同时基团类型“相似性”的提出,改进了随机场基团定义的苛刻性,使得基团类型具有更好的包容性与多样性,可广泛地应用到复杂环境下的多值图像分割问题中。最后给出了该算法的EM参数估计方法和图像分割仿真实验。
Fuzzy Markov random field multi-valued image segmentation using multi-segment fuzzy method to multi-valued problems into a number of binary fuzzy problems, this method is entirely dependent on the pixel ambiguity gray value, which is easy Caught in the local optimal solution. The fuzzy random field based on probability measure gets rid of the dependency of the degree of membership on the gray level, making it easier for the segmentation results to converge to the global optimum. At the same time, the proposed group type “Similarity ” improves the rigor of definition of random field groups, which makes the group types have better inclusiveness and diversity and can be widely applied to multi-valued images in complex environment In segmentation problem. Finally, the EM parameter estimation method and image segmentation simulation experiment are given.